Food & Farming · Wednesday, 22 July 2026
01 · Briefing · what happened
Washington bets AI can save the farm — farmers say their problem is the price of everything
A bipartisan bill would push artificial intelligence onto US farms as the fix for a deepening income crisis. But a fresh survey finds most farmers see no benefit yet, and name the real squeeze themselves: input costs and crop prices set far from the farm gate. Meanwhile drought tightens across Europe and the Horn of Africa.
Key takeaways
- A bipartisan bill would push AI onto US farms to fix a deepening income crisis — but a fresh survey finds 52% of farmers see no benefit yet, and name their real problem as costs and crop prices, not decisions.
- The Farm Bureau projects a sixth straight losing year for major crops, with $32 billion in projected 2027 losses; tractor sales fell 18% as farmers hunker down.
- Drought is tightening across England, central Europe (the Danube at a 30-year low), and the Horn of Africa, where the same cost pressures land with no safety net.
The fix on offer is software. The wound is arithmetic.
On Thursday, a bipartisan bill landed in the House to push artificial intelligence onto American farms. The Fostering Agricultural Research and Modernization through AI Act — the FARM AI Act — comes from Reps. Don Davis, a North Carolina Democrat, and Zach Nunn, an Iowa Republican, with a matching Senate bill from Sens. Ted Budd and Adam Schiff
The pitch is clean. “AI gives farmers better information to make better decisions in real time,” Nunn said
The trouble is what farmers say when you ask them. In the latest Purdue University barometer — a monthly survey of 400 US producers — 52% said AI and data tools give their operation “no meaningful benefit” right now
That is the mechanism worth holding. A commodity crop — corn, wheat, soybeans, the interchangeable bulk goods priced on global markets, where one country’s harvest is measured against everyone’s — earns whatever the world market pays, and costs whatever fertilizer and fuel cost. Both numbers are set far outside any single field. Better decisions inside the field don’t move them.
The losses the software can’t reach
The scale of that squeeze arrived the same week. The American Farm Bureau Federation projects 2027 will be the sixth straight year most major row crops earn negative returns over their total costs
You can see farmers bracing. US tractor sales fell 18% in June against a year earlier, and were down 14% for the year through June
None of that is a decision an algorithm can optimize away. As Michael Langemeier, who runs the Purdue survey, put it: a tool might advise planting corn on a certain day, but weather or a broken machine can make that impossible — the recommendation meets “facts on the ground”
Water is the other bill coming due
Across the Atlantic, the constraint is physical. England and Wales are sliding toward official drought status, and Thames Water, Britain’s largest water company, has joined a growing hosepipe ban
The damage is already priced. An analysis estimates last month’s European heatwave wiped €2 billion from the value of the grain crop, with France and Hungary hardest hit on maize and wheat
Around the beat
Grain markets stayed jumpy but not directional: soybeans rose 21 cents on Monday as China booked 264,000 tonnes for next season, then slipped back on Tuesday
On the science side, one analysis argued cultivated meat — grown from animal cells rather than slaughtered livestock — is edging toward price parity with the conventional kind, though it remains a claim about cost curves, not a shelf reality yet
The cyclospora outbreak that pulled lettuce from Taco Bell last week took an odd turn: the FDA said a key positive test on a Taylor Farms sample was a false positive, even as it kept the supplier as its lead suspect and the recall stayed live across 27 states
The same squeeze, without the cushion
End where the pressure bites hardest. Researchers warn Africa’s Greater Horn — Somalia, eastern Kenya and Ethiopia, South Sudan, Sudan — is heading into what they call a “polycrisis”: below-average rains for a second season, conflict, and the same fertilizer and food-price spikes from the Iran war, all landing at once
It is the same arithmetic that squeezes an Iowa corn farmer — global input costs, global prices, weather no one controls — only here there is no crop-insurance program or emergency-aid debate to soften it. The forces are the same. The cushion is not.
02 · Lesson · why it matters
When the fix assumes the problem is you
A tool aimed at your choices carries a quiet claim — that the limit is inside you. Often it was set for you, somewhere you'll never see.
Two answers to two different problems
Watch what happened this week, side by side. Congress offered American farmers a fix for their deepening losses: artificial intelligence, to help them “make better decisions in real time.” And the farmers, asked to name their own biggest problem, said something else entirely. Not decisions. Costs. Prices. The number on the seed bag and the number the crop sells for.
Both sides are describing the same crisis. But they are answering two different questions. One asks: how do I choose better? The other asks: how do I survive terms I didn’t set? Those are not the same problem, and they do not have the same fix.
A choice problem versus a terms problem
Some problems live inside your choices. You picked the wrong planting date, the wrong field, the wrong moment to sell. A better choice exists, and better information helps you find it. Call it a choice problem.
Other problems live in the terms you choose under. Fertilizer costs what a handful of makers charge. Corn sells for what the world market pays. Both numbers are set before the farmer touches anything, and every choice available loses money against them. The Farm Bureau projects a sixth straight year of this — most major crops earning back less than they cost to grow, corn losses widening from $131 an acre to a projected $167. No planting date fixes a negative equation. That is a terms problem.
Better information is a real gift when the problem is a choice. It is almost nothing when the problem is the terms. You cannot out-decide a price.
The hidden claim inside a fix
Here is why the difference matters beyond farming. Every fix carries an assumption about where the problem lives — and the assumption usually stays silent.
Offer someone a tool to sharpen their decisions, and you have quietly said: the limit is in you. Your choices, your information, your effort. When that’s true, the tool helps and the person feels capable. When it’s false — when the real limit is the terms — the same tool still lands as a verdict. It whispers that a structural squeeze is a personal failing. The farmer running perfect software into a losing market doesn’t just stay broke. He’s handed a story in which the losing was his to prevent.
The researcher who runs the Purdue survey said it plainly: a tool can recommend planting corn on a certain day, but weather or a broken machine can make that impossible. The recommendation meets “facts on the ground.” The facts on the ground were never a decision.
Who set the terms
So look at the shape, not the choice. Beneath the farmer’s losing year sits a structure someone built. Input prices are set by a few large suppliers. Crop prices are set on global commodity markets — where interchangeable bulk goods are priced against every other country’s harvest, so a bumper crop in Brazil can sink the number an Iowa farmer gets, with no one deciding to hurt him.
That structure poses as “just the market” — natural, fixed, nobody’s fault. But it is an arrangement, and arrangements have shapes and beneficiaries. This one keeps food cheap: the same squeeze that drains the farmer fills your grocery cart for less. It can serve the eaters at the top and grind the growers at the bottom in the very same motion. Naming that isn’t blame. It’s seeing the whole thing instead of the farmer’s slice of it.
You are already inside this
The pattern doesn’t stay on the farm. It reaches anyone sold a personal fix for a structural squeeze.
The worker told to reskill, when wages are set by a labor market. The household told to budget harder, when rent is set by how many homes got built. In each case a real, sometimes useful tool arrives wearing the same silent claim: the limit is in you. Sometimes it is. Often the terms were set upstream, in a room you were never in.
And you are not only a witness to the farmer’s version. You are inside it. The cheap food that rests on his losing year is on your plate. Far downstream, in Africa’s Greater Horn, the identical arithmetic — global input costs, global prices, weather no one controls — meets drought and conflict with no crop insurance to soften it, and becomes a hunger crisis. Same forces. Thinner cushion. The web binds the Iowa combine, your dinner, and a failed rain in Somalia into one system, and none of the three can see the other two from where they stand.
The humble part
Seeing the shape doesn’t tell you what to do. AI genuinely helps where the limit really is a decision — the tool is not the villain, and the bill funding it may do real good. The trap isn’t the fix. It’s the misdiagnosis: reaching for a choice-shaped answer to a terms-shaped problem, and mistaking the mismatch for someone’s failure.
The quiet discipline is to ask, before you accept any fix — for a farmer, a worker, a country, yourself — which kind of problem this actually is. Is there a better choice here that I’m missing? Or am I being handed a sharper map of a room whose walls were set long ago, by hands I’ll never see? The honest answer is usually harder to sit with than the tool. It’s also the only one that fits the problem.
03 · Lab · your turn
Decide or Change the Terms
Rehearse a farmer maxing every decision AI can sharpen, discovering it never reaches profit, then feeling the real lever was the terms set upstream.
04 · Hope · carry this
The most hopeful thing this week was quiet: the farmers, closest to the ground, were the clearest of anyone about what actually ails them. Nothing gets mended until it is seen truly — and this week, the seeing was honest.
More from Food & Farming